Closed mysterybear closed 3 years ago
hey @mysterybear can you try running ktunnel with the verbose flag to see if we can catch additional details from the logs?
Hi @omrikiei I was running with the verbose flag already!
I tried strace
, it's pretty noisy, nothing obvious jumping out (was hoping that would hang so we could see where but instead it just keeps logging and logging)
next thing I could try is running the client and server aspects separately rather than using the expose shorthand I guess? I'll try that, pretty busy and did get a tunnel working with frp for my use case this time
Hey @mysterybear, on your client logs, do you see the following log message?
Exposed service's cluster ip is: x.x.x.x
I think this is happening because of a time offset between your machine and the kubernetes node, I will start working on a fix - if you could verify it that would be great!
Hey @mysterybear, on your client logs, do you see the following log message?
Exposed service's cluster ip is: x.x.x.x
Yes that is there, I just didn't include it, sorry for that!
I think this is happening because of a time offset between your machine and the kubernetes node, I will start working on a fix - if you could verify it that would be great!
There's a 1 hour time difference between me:
Wed 14 Apr 15:33:45 BST 2021
And the node:
Wed Apr 14 14:33:21 UTC 2021
Hi,
jfyi: we were hit by the same issue reproduced on mac with timezone difference with v1.3.4, after updating to v1.3.7 the issue was not reproduced.
Note: on ubuntu with v1.3.3 the issue is not present.
Hope that helps, Cheers for the good work with ktunnel! 👏🙇
Hey @mysterybear, issue is resolved on v1.3.7, could you kindly confirm it works now?
Hey @mysterybear, issue is resolved on v1.3.7, could you kindly confirm it works now?
Yes, confirmed! Thanks again @omrikiei !
Image is v1.3.6
Kubernetes side, deployment/pod seems to start up just fine:
Edit: just tried with v1.3.3 also since I remember that working for sure last time I used this (It's me from #36 :D) issue persists just the same... I've checked I haven't got any lingering ktunnel resources server-side as well
Edit:
k logs
on the server pod shows: